{
  "architectures": [
    "MiniMaxForCausalLM"
  ],
  "attention_dropout": 0.0,
  "attn_type_list": [
    0,
    1
  ],
  "block_size": 256,
  "head_dim": 32,
  "hidden_act": "silu",
  "hidden_size": 1024,
  "initializer_range": 0.02,
  "intermediate_size": 3584,
  "layernorm_full_attention_alpha": 1,
  "layernorm_full_attention_beta": 1,
  "layernorm_linear_attention_alpha": 1,
  "layernorm_linear_attention_beta": 1,
  "layernorm_mlp_alpha": 1,
  "layernorm_mlp_beta": 1,
  "max_position_embeddings": 131072,
  "model_type": "minimax",
  "num_attention_heads": 32,
  "num_experts_per_tok": 2,
  "num_hidden_layers": 2,
  "num_key_value_heads": 8,
  "num_local_experts": 8,
  "output_router_logits": false,
  "postnorm": false,
  "rms_norm_eps": 1e-05,
  "rope_theta": 1000000.0,
  "router_aux_loss_coef": 0.001,
  "router_jitter_noise": 0.0,
  "sliding_window": null,
  "tie_word_embeddings": false,
  "torch_dtype": "float16",
  "transformers_version": "4.49.0.dev0",
  "use_cache": true,
  "vocab_size": 32000
}